Why such high Page Rank?

[millington]

Visiting the website www.ncac.com, I was surprised to see it has a high Page Rank of 7, with only 37 backward links. And the links were from pretty ordinary sites, the linking pages having Page Ranks of typically 2 to 4.

Any idea why this site has such a high Page Rank, with only 37 backward links?

[Marcia]

Google only shows a small random sampling of backlinks. Here's a more thorough count - there are a lot more:

http://search.yahoo.com/search?ei=utf-8&fr=slv1-&p=linkdomain%3ancac.com

For starters, here's a quality PR7 link right off. Add all the other links on and it all adds up

http://asa.aip.org/links.soc.html

[millington]

Thank you Marcia. If one wants to find out the actual number of backlinks for a URL, is the best way to go to Yahoo and put linkdomain:URL into the search window? I had not realised Google only gives a selection of backlinks.

[Janet]

I know, my site only shows 7 backlinks on google, but 146 on yahoo. However, yahoo doesn't even show them all since I know of lots that aren't listed there either. Is there any backlink checker that will show ALL you backlinks? I have not found one yet.

[mcanerin]

Keep in mind that a search engine does not, and can not, show you all your backlinks.

All it can show you, at best, is the backlinks it knows about. If that happens to match the actual number of backlinks you have, then, frankly, you are in the minority, and may want to start doing more link building - your link building should always be ahead of a search engines link finding, IMO.

Trivia: Google was started as "backrub", because the founders wanted to do exactly what you want - to find out who was linking to them. In order to do that, they needed to index the web and the links between the sites. Thus Google started.

It's ironic that Google is now the search engine LEAST likely to answer the question for which it was originally created to answer in the first place!

As for Googles reluctance to report backlinks, Danny, myself and several others here are SEW took them to task and managed to convince them to at least stop lying about it. (http://forums.searchenginewatch.com/showthread.php?t=2423)

But they didn't fix it - now they just admit it's broken.... :rolleyes:

Ian

[Janet]

Ian, thanks for the thorough response, very interesting reading. :)

[mphung]

And don't get too caught up in the PageRank as it is displayed in the toolbar. More and more I'm getting the impression that some Google engineers are just playing a giant practical joke on us.

Some of the data centers are giving my personal blog a TBPR of 5/10, which is just ludicrous to me. Even if there are twice or three times as many unique sites linking to it than I'm aware of, that still wouldn't justify a PR5. I.e., my own toolbar PR is grossly inflated ... but that hasn't affected my rankings at all.

I'd focus more on the thematic relevance of the sites than their TBPR when trying to determine how much link value they have.

[Beginner]

Popularity is also an issue. The more people visit a page these days then the more quickly it'll gain a medium high PageRank.

[Marcia]

Popularity is also an issue. The more people visit a page these days then the more quickly it'll gain a medium high PageRank.Nope, not so! PR is still *mathematically* based on linkage, like it's always been.

Usage statistics may figure into scoring for sites, but not where PageRank is concerned.

[Beginner]

Nope, not so! PR is still *mathematically* based on linkage, like it's always been.

Usage statistics may figure into scoring for sites, but not where PageRank is concerned.

I know that my comment is radically different from what's popularly believed but Marcia I trust we can agree to disagree. I'm right. :cool:

[Marcia]

I'm rightAnd how about the source? Any documents out there backing up such a theory? :)

[Beginner]

And how about the source? Any documents out there backing up such a theory? :)

Oh gosh no. Don't make me the first one ever in SEO forum history to have to back up a wild claim. :) Anything I write here is my own personal opinion. I may have research which leads to me form this opinion but in good SEO tradition I get to express my opinion without having to share the research.

If we were all to become scientists and engineers then there would be no room left for webmasters who make pages look pretty with Dreamweaver and free-thinking marketeers! :eek:

[dannysullivan]

Yep, you get to share your opinion. If you can't back it up, it remains that, an opinion.

I've never seen Google suggest that PageRank is influenced by visits. Nada.

You're the first person I've ever heard suggesting the connection, at all.

I have heard suggestions that search engines might take visitation into account when assigning credit on how to weight web sites. Google seems to be clickthrough tracking everything these days, plus they've have visitation data that flows in from the toolbar. It wouldn't suprise me at all if these metrics are being used to influence rankings.

That's not exactly the same as being used to make the little PageRank meter show more, of course. Nor is it necessarily the same as adding to PageRank. PageRank might be one factor that remains based on counting links, while another PopularityRank factor, if you will, could be based on looking at visitation data.

[Beginner]

You're the first person I've ever heard suggesting the connection, at all.


Excellent! I'm going to bookmark this thread and hope to be able to cite it for bragging rights later on. :cool:

After being so open about how links influence their algorithm and seeing the huge rise of unnatural links that that caused I do suspect Google will be a lot more coy about sharing their current implementation.

That said, their 'historical data' patent application does hint at this.

[Marcia]

Nah, even before that historical data patent was issued - there was another granted back in 2002, when click tracking had already been observed for a while:

Methods and apparatus for employing usage statistics in document retrieval (http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220020123988%22.PGNR.&OS=DN/20020123988&RS=DN/20020123988)

Yep, you get to share your opinion. If you can't back it up, it remains that, an opinion.Which underlines the wisdom of keeping in mind, when reading unsubstantiated speculation in forum posts:

The Law of Logical Argument:
Anything is possible if you don't know what you are talking about.

[Mel]

Marcia

Do you mind if I borrow that quote from time to time (with proper attribution of course)

[Marcia]

Borrow away Mel, gratis. It originally came from one of those joke things that makes its way around the web that has a whole list of "laws" that are "only too true."

[nima1353]

Food for thought....

I was wondering what your interpretation of the following claim in Google patent is:

Quoted:
34. The method of claim 1, wherein the one or more types of history data includes information relating to traffic associated with documents; and wherein the generating a score includes: determining characteristics of traffic associated with the document, and scoring the document based, at least in part, on the characteristics of traffic associated with the document.
Unquoted.

Could it mean they might be watching your traffic level?

Nima

[Marcia]

nima, sure it's entirely possible that traffic stats are used in scoring - they've just got nothing to do with computing PageRank.

[Robert_Charlton]

nima, sure it's entirely possible that traffic stats are used in scoring - they've just got nothing to do with computing PageRank.

Marcia - Of course I agree. I'm thinking I need to say this before going on, because some wildly off-base opinions have been expressed here.

I've been thinking as I read the thread that it's possible that several of the posters don't know what PageRank is, and that it's different from rankings in the serps.

It's also possible that they don't know the difference between correlation and causality....

It's very likely, for example, that if a site is a good enough, more traffic will bring about an increase in the number of inbound links to the site, and that would in turn increase PageRank. The increase in traffic would correlate with the increase in PageRank, but the traffic stats themselves would not cause the increase in PageRank.

The number of visitors and the computation of PageRank are independent of each other. The end effects of the links that site visitors happen to direct to the site would be what causes the increase.

[Beginner]

Urg. Pride is forcing me to post again - apologies if it seems that I can't let the thread rest.

I know what PageRank is. I know how it's calculated - it's calculated as described above; with links, where not every link is equal.

My claim - and it is a claim, since I'm not willing to share suitable details of my research - is that traffic popularity effects the rate at which a high/medium PageRank can be attained.

For example, you manage to get some great inbound links, some from wikipedia, some from local .gov sites, some from classic directories like Yahoo and Business.com. If no one visits your site, if you have no market share, then you'll not gain that potential PageRank as quickly as you could. Think of traffic popularity as a throttle.

Again, to stress, this is something that my research stresses but as almost everything else you read on these forums this is not sacrosanct fact. It's my speculation but it just so happens to be based on research I find very convincing.

I'm not prepared to name explicit test sites. If I did then it'll skew the test. However, imagine an email campaign which links to two new pages on a web page. The call to action on one link is strong. The call to action on the other is weak. Traffic logs show lots of people going to the 'strong' page.

In this scenario the strong page quickly romped to PageRank 6. The other, the weak page, sits still at 0.

Of course, the strong page is more likely to attract natural links as people talk about it. There's little evidence of this based on Yahoo's and MSN's link commands (quicker, less fussy). In fact, as this is not a phase one experiment, there are deliberately high quality links (including one .gov source) to the 'poor call to action' page. These inbounds are shown on Yahoo and MSN. As with the other experiments, there's no sign of referral traffic direct to these web pages either.

Another experiment works with an imagemap. One hot spot is attractive and gains most of the traffic. The rest of the hot spots are dull and rarely visited. There is no sign of any inbound links to any of these pages, in fact, http_referrer checks automatically redirect traffic out of the hot spot destination pages unless the user has entered via the imagemap page. It's the high traffic hot spot destination page that has twice the PageRank of the others.

There are other, more complicated, examples but this post is long enough. Also, I'm aware, there are people who'll only want more explicit examples which I'm not willing to give.

I now think of linkStrength to be the tally by which maximum potential PageRank is calculated. Traffic popularity/market share is a filter - and a dominant one - which controls/caps/throttles how much of that potential is actually reached.

[incrediblehelp]

I know there is life on other planets, but I am not going to tell you which ones.

[Mel]

It would seem pretty silly for Google to continue to pay Stanford university large payments every year to license the use of PageRank if they are now using a system which is not Pagerank, i.e, factors in traffic.

I can imagine that the educational level of searchers who click on a link might be a good way to assign some sort of a rank to such clicks, but that certainly does not mean the Google has either the capability or desire to do that or that they are in fact doing it.

If someone has actually come up with the Rosetta stone to the Google algo, I will be the first to cheer him/her but they will have to show some examples first.

Sorry not even the smallest amount of evidence that this or that may be happening means that I discard that theory without further ado.

[nima1353]

PageRank as we knew it.....

A brilliant idea that helped Google frog leap its competitors......then came the link building rush....more and more webmasters started to add or buy back links to their sites....to get higher PR....a direct attack to the heart of algorithm to manipulate its results!!!

So, even though PageRank was a great idea at first, it had its shortcomings. Google needs to come up with more sophisticated and/or additional factors to augment its algorithm.

One of the factors can be historic data on website traffic. It might not be used as a variable in the calculation of PageRank as we knew it, but it can certainly be used in conjunction with PR to rank sites.....the same way GEO targeting has an affect….

Increase in traffic can be contributed to many factors such as offline advertising, email marketing, PPC advertising and so and so forth. So, it is not a straight forward correlation.

[Mel]

IMO PageRank has not been a significant Google ranking factor for several years now, but anchor text links are used as a significant ranking factor.

The idea of using any type of traffic stats as an important ranking factor is far fetched IMO as the top ranking sites always get the most traffic which would in turn boost thier rankings, which would in turn generate more traffic.....

[Wail]

IMO PageRank has not been a significant Google ranking factor for several years now, but anchor text links are used as a significant ranking factor.

The idea of using any type of traffic stats as an important ranking factor is far fetched IMO as the top ranking sites always get the most traffic which would in turn boost thier rankings, which would in turn generate more traffic.....

There is strong evidence (in the form of patent applications) (but I suppose that doesn't mean Google is using that technology; just stopping others from doing so) that traffic can be a ranking factor. Whether that means traffic is also a PageRank factor, I don't know.

A site at position #1 for a popular keyword would be expected by Google to get x% of all the search hits. A site at position #2 for the same keyword would be expected to get y% of all the search hits.

If the site at position #1 actually only manages to get x-5% of all the search hits then I can see why Google would use that as an indication that the site did not seem to be matching the users search expectations and I would imagine that that site would begin to drift down the SERPs.

Of course, this is speculation. I have to point out that much of canon SEO is speculation. We take it for granted that H tags embiggen (http://www.mattcutts.com/blog/seo-mistakes-unwise-comments/) keywords and that title tags are very important too. Sure, we all see the effects these changes have, but nothing we do is done in isolation. Change your title tag and you'll change your snippet in SERPs. If my speculation about traffic is correct then that alone can effect your rankings then perhaps that would be responsible for the change. Though I don't think so.

I've found this thread interesting. If someone speculates 'B' then sometimes the forums can throw up their hands in praise. If someone else speculates 'C' then sometimes the forums can squish them down. The suggestion that traffic might be important tends to be unpopular and therefore, I see, a reluctance to believe it.

I see room for an online experiment in anthropology.

[Mel]

There is also evidence from Patent applications that the number of years you have paid in advance for your domain name is a ranking factor too, but as you said actually using these strange ideas may not have been the reason for the patent.

There are two major factors to getting traffic from SERPs - ranking well and having a snippet which entices viewers to click thru so if you want to use traffic as a ranking factor you are going to have to develop a method of evaluating the clickability of each individual SERP title and description.

So lets suppose that all these things are taken care of and traffic is uses as you propose ---> what this means is that new sites don't stand a chance since they can't get top rankings without search engine traffic and they can't get search engine traffic because they can't rank - Catch 99.

So then some clever fellow develops a bot which, for a slice of your income, will click on your serps for you as many times as you like and so the cycle continues.

[chrisnocom]

My question is:
If the link rapport for my website in Google only shows one link, does that mean that Google's robots only have found one page linking to mine and thus will not increase my PR? (you say the search engine only show the backlinks it knows about).

I find it strange that Google only shows one backlink as Yahoo and Altavista for a couple of months now have found more backlinks (allthough not all of them).

Is it also so that the moment, let's say Google find 5 sites with a PR 4 linking to my site, my PR will increase, or will it also take a while from the baclinks are found to the day the PR increases?

Thanks for answers!

Christian Bromander

Keep in mind that a search engine does not, and can not, show you all your backlinks.

All it can show you, at best, is the backlinks it knows about. If that happens to match the actual number of backlinks you have, then, frankly, you are in the minority, and may want to start doing more link building - your link building should always be ahead of a search engines link finding, IMO.

Trivia: Google was started as "backrub", because the founders wanted to do exactly what you want - to find out who was linking to them. In order to do that, they needed to index the web and the links between the sites. Thus Google started.

It's ironic that Google is now the search engine LEAST likely to answer the question for which it was originally created to answer in the first place!

As for Googles reluctance to report backlinks, Danny, myself and several others here are SEW took them to task and managed to convince them to at least stop lying about it. (http://forums.searchenginewatch.com/showthread.php?t=2423)

But they didn't fix it - now they just admit it's broken.... :rolleyes:

Ian

[attak]

To think that Google divulges ALL of it's information regarding how they calculate PR (or anything else) would be naive at best.
"Beginner" is correct at some level with his statement regarding traffic/popularity being a PageRank enhancer...

At the very least you could consider this possible scenario:
Site A has 1,000 IBLs with mediocre traffic averages.
Suddenly Site A has a jump in it's natural and "link" traffic that triples the amount it used to average over the course of the previous months. For the sake of argument let's say that no additional IBLs are gained.

If this traffic is sustained over a period of a couple months isn't it feasible that Google assigns Site A higher TrustRank - or views Site A as more of an Authority Site than previously - which would in turn give their IBLs (link campaigns) more credibility/relevancy/trust/weight and result in an increase in Site A's PR?

I believe so...and this isn't the only scenario I could see this happening with. Then again, maybe I'm crazy for thinking a site like Apple.com would still have an extremely high PR without a single IBL :rolleyes:

Attak

[sakib000]

Visiting the website www.ncac.com, I was surprised to see it has a high Page Rank of 7, with only 37 backward links. And the links were from pretty ordinary sites, the linking pages having Page Ranks of typically 2 to 4.

Any idea why this site has such a high Page Rank, with only 37 backward links?

how is that

[JohnW]

>how is that

Please read the thread. The answer has been given.